Accurate cell type deconvolution in spatial transcriptomics using a batch effect-free strategy
Wang, L.; Wu, L.; Liu, C.; Wang, W.; Zhang, X. H.- F.; Liu, Z.
Show abstract
Sequencing-based spatial transcriptomics (ST) techniques have been groundbreaking in dissecting cell-cell communications within tissues by profiling positional gene expression. However, the most widely used ST technique, Visium Spatial Gene Expression by 10x Genomics (Visium), does not provide single-cell resolution, making it difficult to profile cell type-level information. Many reference-based deconvolution methods have been developed to increase its resolution, but the platform and batch effects between the reference and ST data compromise their accuracy. Here, we propose a new approach, Region-based cell Sorting (ReSort), that generates a pseudo-internal-reference to reduce these platform effects. By simulating ST datasets under various scenarios, we demonstrate that ReSort significantly improves the accuracy of six state-of-the-art reference-based deconvolution methods. Moreover, applying ReSort to a mouse breast cancer tumor bearing both epithelial and mesenchymal clones identifies the spatial differences of immune cells between the clones, providing important insights for understanding the relationship between epithelial-mesenchymal transition and immune infiltration in breast cancer.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Spatially informed clustering, integration, and deconvolution of spatial transcriptomics with GraphST 97%
- stLearn: integrating spatial location, tissue morphology and gene expression to find cell types, cell-cell interactions and spatial trajectories within undissociated tissues 97%
- ROGUE: an entropy-based universal metric for assessing the purity of single cell population 97%
Similar papers in this journal
- High-precision cell-type mapping and annotation of single-cell spatial transcriptomics with STAMapper 97%
- stGCL: A versatile cross-modality fusion method based on multi-modal graph contrastive learning for spatial transcriptomics 97%
- Explainable multi-view framework for dissecting inter-cellular signaling from highly multiplexed spatial data 96%
Similar papers in this journal
Similar papers in this journal
- Clustering-independent estimation of cell abundances in bulk tissues using single-cell RNA-seq data 97%
- Multi-modal image analysis for large scale cancer tissue studies within IMMUcan 96%
- SELINA: Single-cell Assignment using Multiple-Adversarial Domain Adaptation Network with Large-scale References 96%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.